Al2Mo3O12/polyethylene composites with reduced coefficient ... · Al2Mo3O12/polyethylene composites...

of 29/29
Corresponding author. Tel: +55-21-35271954; fax: +55-21-35271236 E-mail address: [email protected] (B.A. Marinkovic) Al 2 Mo 3 O 12 /polyethylene composites with reduced coefficient of thermal expansion Alexandre Roberto Soares, Patricia I. Pontón, ψ Lidija Mancic, § José R. M. d'Almeida, ψ Carl P. Romao, л, Mary Anne White, л, ,ξ Bojan A. Marinkovic , ψ * Centro Universitário de Volta Redonda – UNIFOA, Av. Paulo Erlei Alves Abrantes 1325, Volta Redonda – RJ, Brasil ψ Departamento de Engenharia de Materiais, Pontifícia Universidade Católica de Rio de Janeiro – PUC-Rio, Rua Marquês de São Vicente 225, Gávea, Rio de Janeiro, RJ, Brasil § Institute of Technical Sciences of SASA, Knez Mihailova 35/IV, Belgrade, Serbia л Department of Chemistry, Dalhousie University, Halifax, Nova Scotia, B3H 4R2, Canada Institute for Research in Materials, Dalhousie University, Halifax, Nova Scotia, B3H 4R2, Canada ξ Department of Physics and Atmospheric Sciences, Dalhousie University, Halifax, Nova Scotia, B3H 4R2, Canada e-mail addresses : [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] [email protected] *
  • date post

    20-Jan-2019
  • Category

    Documents

  • view

    218
  • download

    0

Embed Size (px)

Transcript of Al2Mo3O12/polyethylene composites with reduced coefficient ... · Al2Mo3O12/polyethylene composites...

Corresponding author. Tel: +55-21-35271954; fax: +55-21-35271236 E-mail address: [email protected] (B.A. Marinkovic)

Al2Mo3O12/polyethylene composites with reduced coefficient of thermal expansion

Alexandre Roberto Soares,Patricia I. Pontn,Lidija Mancic, Jos R. M. d'Almeida,

Carl P. Romao,, Mary Anne White,, , Bojan A. Marinkovic, *

Centro Universitrio de Volta Redonda UNIFOA, Av. Paulo Erlei Alves Abrantes

1325, Volta Redonda RJ, Brasil

Departamento de Engenharia de Materiais, Pontifcia Universidade Catlica de Rio de

Janeiro PUC-Rio, Rua Marqus de So Vicente 225, Gvea, Rio de Janeiro, RJ, Brasil

Institute of Technical Sciences of SASA, Knez Mihailova 35/IV, Belgrade, Serbia

Department of Chemistry, Dalhousie University, Halifax, Nova Scotia, B3H 4R2,

Canada

Institute for Research in Materials, Dalhousie University, Halifax, Nova Scotia, B3H

4R2, Canada

Department of Physics and Atmospheric Sciences, Dalhousie University, Halifax,

Nova Scotia, B3H 4R2, Canada

e-mail addresses:

[email protected]

[email protected]

[email protected]

[email protected]

[email protected]

[email protected]

[email protected]*

2

Abstract

Recently, polymer composites reinforced with lowfractions of thermomiotic

nanoceramics have triggered a lot of research. The efforts have been focused on

achieving considerable reduction of the coefficient of thermal expansion (CTE) of

polymeric materials without deterioration of other physical properties. In this context,

polyethylene (PE) composites reinforced with different loads of Al2Mo3O12 nanofillers

(0.5 to 4 mass %) were fabricated by micro-compounding. To enhance the

interfacialinteraction between the two components chemical functionalization of

Al2Mo3O12 was performed with vinyltrimethoxysilane (VTMS) prior to micro-

compounding. Infrared spectroscopy and thermogravimetry demonstrated the successful

grafting of VTMS on the Al2Mo3O12 surface. The composites showed strongly

decreased CTEs, up to 46 % reduction for loadings of 4 mass % compared with neat PE,

suggestingintimate filler-matrix interactions. The variation of CTEs of the composites in

terms ofthe fillerfraction was successfully described by Turners model

allowingcalculation ofthe bulk modulus of monoclinic Al2Mo3O12 (13.6 2.6 GPa),in

agreement with the value obtained by an ultrasonic method. The thermal stability of

thecomposites was improved, although the addition of functionalized fillers decreased

the degree of crystallinity of the PE to a small extent. The Youngs modulus and yield

strength of the compositesincreased from6.6 to 19.1 % and 4.0 to 6.0 %, respectively,

supporting the existence of strong filler-matrix interactions, contributing to an efficient

load transfer. Finite element analysis of thermal stresses indicated absence of plastic

deformation of the matrix or fracture of thenanofillers, for a 100 K temperature drop.

Keywords: negative thermal expansion,ceramics, bulk modulus, thermoplastics,

chemical functionalization, nanomaterials.

3

1. Introduction

Negative thermal expansion (NTE) ceramics, also known as thermomiotics[1],

began to attract significant attention in the 1990s due to newly discovered phases

exhibiting this uncommon thermal expansion feature [2-4]. Since then, considerable

research has been carried out to advance understanding of the physics underlying this

phenomenon [5-8], while parallel efforts have been made to transform this new class of

compounds into useful materials. There are two general tentative approaches to engineer

new materials based on thermomiotics. One route corresponds to the development of

bulk monolithic ceramics with well-defined CTEs, generally free of undesired

microstructural defects (microcracks) and frequently aiming at near zero values [9].

However, in some cases, such as Al2TiO5, microstructural design and microcrack

formationcan be used to advantage to produce low thermal expansion materials with

high thermal shock resistance [10]. On the other hand, the second approach makes use

of the addition of thermomiotics, as the dispersed phase, in composites [11-29]. The

most common role of thermomiotics in composites is to control the CTE of the matrix

phase, which could be a metal, ceramic or polymer.

Thermoplastic polymers generally present very poor dimensional stability on

temperature variation. Therefore, the addition of inorganic fillers with low positive, near

zero or negative CTEs can mitigate high CTEs in polymers, while direct or silane

crosslinking strategies are also employed in order to improve their physical properties.

An example of the importance of controlling thermal expansionin polymers is observed

in the field of electronic packaging polymers, such as polystyrene [30]and thermoset

epoxy resins [31], applied over integrated circuit chips to protect them from

environmental factors. These polymers are usually filled with a ceramic showing low

CTE and moderate thermal conductivity for the sake of reduction of thermal mismatch

between the packaging polymer and the substrate, while also increasing the thermal

conductance of the package layer. Polyethylene (PE) is one of the most used

thermoplastics. Geomembranes and pipes for water or natural gas

transmission/distribution are some examples of PE applications where the huge CTE (~

10-4 K-1) or migration of volatile organic components (VOCs) can induce material

damage orlead to poor performance [32, 33]. The list of dispersed phases studied for the

purpose of reduction of CTE of PE is extensive and includes clays, CaCO3, Si, fly ash,

CeO2, Sr2Al2SiO7, just to mention a few [34-39], but, so far, has not included

thermomiotics or low CTE materials.

4

Regarding thermomiotic fillers, there are only a few reports in the literature of

attempts to control the CTE of polymers, mostly thermosetting matrices.

Thesetentativeattemptsgenerally focus on ZrW2O8, both as micronic and nanometric

powders [11, 15-17, 23-25]. This is understandable considering the isotropic physical

properties of ZrW2O8and its strong negative CTE (~ - 9 x 10-6 K-1) over a wide

temperature range (0-1050K), which includes theoperational temperature range for

polymer components. This NTE material has been used for the reduction of CTE of

polyimide, polyester and epoxy resin [11,15-17,23,24], while, as far as the authors are

aware, there are no reports on its use in a PE matrix. On the other hand, the

thermomiotics and low CTE compounds from the A2M3O12 family are rarely cited in the

literature as the dispersed phases for controlling CTE in composites [14, 18, 19, 22,

40].An advantage of the compounds from this family is their wide chemical flexibility,

resulting in a variety of orthorhombic phases with CTEs ranging between 11 x 10-6

and 2 x 10-6 K-1 [8, 41, 42]. Even in the monoclinic phase, the magnitude of the CTE of

A2M3O12 compounds is low [9]. It is also significant that some of the A2M3O12 compounds are routinely synthesized from inexpensive reagents,which is not the case

for ZrW2O8 and related compounds. There are currently just a few reports of the

application of A2M3O12 compounds as fillers in ceramic-ceramic (MoO3, ZrO2, ZrSiO4 and ZrW2O8 featuring as matrices) [14, 19, 22, 40]and ceramic-metal (Mg and Cu

matrices) [18, 40]composites. However, there are no reports on the use of A2M3O12 in

polymer matrices. The A2M3O12 compounds already tested in composites as dispersed

phases are Y2W3O12, Sc2W3O12, FexSc1-xW3O12 and ZrWP2O12, while the application of

Al2Mo3O12 was only reported in the monolithic form [9].

The principal purpose of this study was to examine the viability and potential

employment of a nanometric-sized A2M3O12 as a filler for thermoplastic matrices, added

in low amounts ( 4 mass %), to reduce the CTE of thermoplastic polymers without

provoking adverse effects on their mechanical properties and thermal stability. Because

of its technological appeal, PE has been chosen as the matrix, while Al2Mo3O12, a

reasonably well studied material[9, 43, 44], has been elected as the nanofiller, although

in the temperature range below 473 K it adopts the monoclinic form with a low positive

CTE [9]. To enhance the interface interaction between the inorganic hydrophilic filler

and the hydrophobic polyolefin matrix, chemical functionalization of Al2Mo3O12 nanoparticles was carried out with the aid of a bi-functional coupling agent,

vinyltrimethoxysilane (VTMS). One functional group of VTMS is hydrolysable

5

trimethoxysilane ((Si(OCH3)3)), which is capable of establishing primary bonds with

the surface of the inorganic filler, while the other functional group is a hydrophobic

vinyl group (CH2=CH-), capable of interacting physically (inter-molecular

entanglement) or chemically, through primary bonds created in the presence of peroxide

initiators, with PE chains. The composites were processedby micro-

compounding,includingextrusion and injection molding.

2. Materials and methods

2.1 Materials

Starting reagents for synthesis of Al2Mo3O12 nanopowder and its

functionalization include Al(NO3)3 9H2O (Isofar, 99 %), (NH4)6Mo7O24 4H2O (Isofar,

99 %), NaOH (Vetec, 99 %), vinyltrimethoxysilane C5H12O3Si (Evonik, 98%), acetic

acid (Vetec) and anhydrous ethanol (Sigma Aldrich). All reagents were used asreceived.

Granulated medium density PE (MR-435, Braskem) with a density of 0.934 gcm-3 and

melt flow index (MFI) of 4.0 g/10 min served as the matrix.

The synthesis and characterization of Al2Mo3O12 nanopowder through a co-

precipitation route is described in detail elsewhere[9].

Bulk Al2Mo3O12 pellets were prepared in order to measure their mechanical

properties using ultrasonic transduction. Thesynthesis was performed using a solid-state

reaction method, using Al2O3 (Sigma-Aldrich, 99.7 %) and MoO3 (Sigma-Aldrich, 99.5

%) as precursors. The precursor powders were mixed and mechanically activated in a

high-energy ball mill for 12 hours, followed by consolidation in a cylindrical die press

(15 mm diameter) under approximately 55 MPa of pressure and subsequent reactive

sintering for 24 h at 1050 K in air. The three resulting pellets had an average height of

4.78 mm and a density fraction 73 % of theoretical.

2.2 Functionalization of Al2Mo3O12 nanoparticles with VTMS

The functionalization of Al2Mo3O12 was carried out using 51 mL of

ethanol/water 95:5 % v/v as the reaction medium[45]. A mass of 1g of Al2Mo3O12,

obtained though the co-precipitation route and ground into a fine powder, was added to

aqueous ethanol and stirred for 10 min, while the pH was adjusted with acetic acid to

4.5-5.5. The suspension was placed in an ultrasonic bath for 1 h and the pH was

adjusted again to 4.5-5.5. VTMS was added dropwise until a mass ratio of 1:2 (VTMS:

Al2Mo3O12) was reached. This suspension was mechanically stirred at a frequency of

5000 min-1 (Ultra Turrax Ika T25 disperser) for 2 h, then ultrasonicated for a further 1 h.

The as-grafted Al2Mo3O12 nanoparticles were washed three times with aqueous ethanol

6

to remove the excess VTMS, and then filtered and dried at 383 K for 2 h in an air-

circulating oven. Here the silanized Al2Mo3O12 nanoparticles are referred to as VTMS-

g-Al2Mo3O12, where g denotes that VTMS was grafted onto the nanoparticle

surfaces.

2.3 Preparation of VTMS-g-Al2Mo3O12/PE composites

VTMS-g-Al2Mo3O12/PE composites were prepared in a twin-screw micro-

extruder (DSM Xplore 5cc Microcompounder). The screw speed was 100 rpm, the

barrel temperature profile 443, 448 and 453 K (from the top to the bottom) and the

mixing time was approximately 5 min. The extrudate was fed to a micro-injection

molding machine (DSM Xplore 5.5 cc) to obtain ASTM D638-10 tensile specimens.

The injection pressure was set to 7 bar. The melt and mold temperatures were kept at

448 K and 353 K, respectively. To investigate the effect of different loads of VTMS-g-

Al2Mo3O12 on CTE, thermal stability and mechanical properties of PE,composites

containing nominally0.5, 1.0, 2.0, 3.0 and 4.0 mass % of VTMS-g-Al2Mo3O12 were

prepared and compared with neat PE specimens.

2.4 Characterization

2.4.1 Al2Mo3O12 powder

The Al2Mo3O12 powder obtained through co-precipitation was characterized by

X-ray powder diffraction (XRPD) on a Siemens D5000 using CuK radiation in steps of

0.02o (8 s per step) from 10o to 60o (2 theta) to determine phase composition of the

product and to evaluate the average crystallite size of Al2Mo3O12. Powder Diffraction

File 2 Release 2011(PDF-2) was used for crystal phase identification, while Topas 4.2

was used for LeBail analysis. The direct convolution approach, embedded in Topas 4.2,

has been used in order to separate physical from instrumental contributions within the

experimental pattern. The instrumental contribution has been described through the

Fundamental Parameter Approach (FPA), while the physical (microstructural)

contribution was modelled by two Voigt functions. The reported crystal size parameter

is LVOL-IB obtained from the integral breadth of the Voigt function that describes

crystallite size contribution to the physical broadening. This approach for determination

of crystallite sizes is known as the Balzar method.

Transmission electron microscopy (TEM) was performed on a JEOL-2010

microscope operating at 200 kV using a Gatan CCD camera. TEM specimens were

prepared by dispersing the powder samples in isopropyl alcohol via ultrasonic treatment

and then dropped onto a porous carbon film supported by a copper grid. Powder

7

samples, following pretreatment at 393 K under vacuum (50 mTorr), were characterized

by N2 adsorption at 77 K on a Micromeritics TriStar 3000V6.03 instrument. Specific

surface area was calculated using the BrunauerEmmettTeller (BET) equation in the

P/Po range of 0.060.21.

2.4.2 Bulk Al2Mo3O12

The transverse and longitudinal velocities of sound in three Al2Mo3O12 pellets

were measured ultrasonically using a Panametrics 25D Ultrasonic Thickness Gauge

with Panametrics shear-wave couplant and glycerol (longitudinal-wave couplant). The

measured velocities of sound were corrected for porosity as follows [46]:

, (1)

where is the measured velocity, 0 is the corrected, fully densified, velocity, and p is

the pore fraction.

2.4.3 VTMS-g-Al2Mo3O12

Fourier transform infrared spectroscopy (FTIR) was performed in the range

4000 to 500 cm-1, using pellets with the addition of dried IR-grade KBr,in a Perkin-

Elmer Spectrum 100 FT-IR spectrometer with a resolution of 0.5 cm-1.

Thermogravimetric analyses (TGA) were carried out on a Perkin-Elmer

Simultaneous Thermal Analyzer (STA-6000). About 5 mg of VTMS-g-Al2Mo3O12 was

added to an alumina pan (180 L) and heated in N2 gas flux (50 mLmin-1) in the

temperature range between 298 and 853 K, at a heating rate of 10 Kmin-1. The

thermogravimetric balance was previously calibrated in N2 gas flux (50 mL min-1) in the

range between 298 K and 1173 K using a 15 mg of CaC2O4 H2O standard sample,

supplied by Perkin-Elmer. The buoyancy effect, observed in TG curves (as a small mass

gain, < 0.1 mass %, at temperatures lower than 330 K),was partially corrected by blank

tests. Both analyses, FTIR and TGA, were also performed on Al2Mo3O12 powder before

grafting, for comparison.

2.4.4 VTMS-g-Al2Mo3O12/PE composites

Dilatometric curves of neat PE and VTMS-g-Al2Mo3O12/PE composites were

determined in air atmosphere using a NETZSCH dilatometer DIL 402C in the

temperature range from 298 to 373 K with a heating rate of 10 Kmin-1. The specimens

for dilatometry were prismatic rods (3.3 mm 3.2 mm 11 mm) manufactured from

the injection molded specimens. Each material, neat PE and VTMS-g-Al2Mo3O12/PE

composites, was tested using threeseparate samples. The instrument was assessed for

8

accuracy using an -alumina rod supplied by NETZSCH. The accuracy of L/Lo (thermal strain; Lo is the original sample length) is ~ 1 % for the dilatometer used [47].

TGA and differential scanning calorimetry (DSC) analyses were carried out on

the equipment described in section 2.4.3 using the same conditionsand corrections, but

in the temperature range 298 to 923 K.Simultaneous Thermal Analyzer (STA-6000)

was calibrated for temperature and enthalpy using ~ 5 mg of indium and silver standard

samples, supplied by Perkin-Elmer. TG curves of neat PE and VTMS-g-Al2Mo3O12/PE

composites are reported in the Online Resource.

Tensile tests were carried out on ASTM D638-10 tensile specimens, at room

temperature, on an EMIC Universal Testing Machine (model DL 1000) with the

crosshead speed separation of 30 mm/min. Each material (neat PE and VTMS-g-

Al2Mo3O12/PE composites) was tested using sixseparate specimens.

2.4.5 Finite element analysis

Finite element analysis (FEA) was used to assess the thermal stress that could be

expected in VTMS-g-Al2Mo3O12/PE composites upon cooling. The COMSOL

Multiphysics 4.4 software package was used to perform the analysis. The model

geometry chosen was that of nine spherical filler particles of monoclinic Al2Mo3O12

embedded at random positions in a cube of PE, where the volume fraction of filler was

varied to match the experimental compositions.A mesh of tetrahedral elements was

generated in COMSOL and the mesh size used was shown by a mesh convergence

study to produce results convergent to within 1%.

The VTMS grafting was assumed, based on the experimental results described

below, to produce perfect interfaces between matrix and filler.Both Al2Mo3O12 and PE

were treated as linear elastic materials. Assumption of perfect interfaces and linear

elasticity leads to the calculated stress being a conservative estimate; the stress in the

actual composite could be reduced by imperfect interfaces or ductility. Therefore, the

results of the finite element analysis should be treated as qualitatively accurate.Due to

the low volume fractions of filler used, the size, shape, and distribution of the particles

did not substantially influence the results and these parameters were not varied as part

of this study. The boundary conditions used were a temperature drop of 100 K on all

sides of the material and roller boundaries on three perpendicular sides.

9

3. Results

3.1 Al2Mo3O12 powder

The experimental XRPD pattern (Figure 1) confirmed that the co-precipitation

route resulted in a single-phase powder, since all diffraction lines correspond to

monoclinic Al2Mo3O12 (PDF: 01-089-8579). However, the well-defined diffraction

lines of monoclinic Al2Mo3O12were overlaid on a broader diffraction pattern. A

thorough examination of the powder by TEM revealed an intimate mixture of well-

defined nanometric crystals (50 to 200 nm) with much smaller nanocrystallites(ca. 10

nm); see Figure 2. (A few other TEM images presented in Online Resource illustrate the

presence of larger particles.)This observation was confirmed more quantitatively by

applying LeBail analysis considering two Al2Mo3O12 phases with the same monoclinic

structure but different mean crystal sizes (Figure 1). Only this fitting approach resulted

in good agreement with the experimental pattern, giving average crystal sizes of 70 nm

for the well-defined Al2Mo3O12 and 10 nm for the broader Al2Mo3O12 diffraction

pattern. BET analysis of N2 adsorption indicated a specific surface area of 8.4 m2g-

1.Considering a crystallographic density of 3.495 g cm-3 for monoclinic Al2Mo3O12 [9],

an average (spherical) particle size of ~ 100 nm was calculated from the observed

specific surface area (8.4 m2 g-1), which is somewhat larger than the crystallite sizes

obtained from XRPD data.

Fig.1Experimental XRPD patternof monoclinic Al2Mo3O12 (red dotted line), the pattern

fit bythe Le Bail method (black line) and the difference plot (green line)

10

Fig.2 TEM images of Al2Mo3O12 nanometric powdershowing a mixture of crystal sizes

(a) from 50 to 200 nm(arrows indicate faces of larger particles)and (b) ~10 nm

3.2 Bulk Al2Mo3O12

Following correction for porosity (Eq. 1), the longitudinal velocity of sound of

monoclinic Al2Mo3O12 was determined as 3.29 0.16 m ms1and the transverse

velocity of sound was determinedas 2.29 0.22 m ms1. Using the reported theoretical

density of monoclinic Al2Mo3O12 of 3.495 g cm3[9], these values correspond to a bulk

modulus of 13.5 1.3 GPa and a shear modulus of 18.2 0.7 GPa. Interestingly, this

indicates that monoclinic Al2Mo3O12 has a near-zero Poisson ratio, specifically 0.036

0.088.

3.3 VTMS-g-Al2Mo3O12

FTIR spectra of pristine Al2Mo3O12 and VTMS-g-Al2Mo3O12 are presented for

comparison in Figure 3.

11

Fig.3 FTIR spectra of (a) pristine Al2Mo3O12 and (b) VTMS-g-Al2Mo3O12

The spectrum of VTMS-g-Al2Mo3O12 shows four absorption bands at 2920,

2855, 1262 and 1154 cm-1 that suggest grafting of VTMS onto Al2Mo3O12. The bands at

2920 and 2855 cm-1 are generally assigned to stretching vibration modes () of CH3 in

(OC2H5) and CH2 in (CH=CH2) [48], respectively, while the band at 2920 cm-1 is also

attributed to stretching vibration modes of CH3 in (OCH3) [49, 50]. However, these two

bands at 2920 and 2855 cm-1are also faintly observable in the spectrum of pristine

Al2Mo3O12 and are attributed to organic surface contamination [51]. The band at 1262

cm-1 appears only in the VTMS-g-Al2Mo3O12 spectrum and is assigned to the in-plane

CH bendingmode () in (CH=CH2) [48]. In addition, there is a band at 1154 cm-1

attributed to (Si-O-Si) [52], proving the occurrence of oligomerization between silanol

reaction groups grafted on the Al2Mo3O12 nanoparticles.

From the TGA results, a total mass loss of ~ 1.1 mass% of pristine Al2Mo3O12

from room temperature up to 853 K (Figure 4) was observed and is attributed to

physically adsorbed water (moisture) and primary bonded hydroxyl groups[53].

12

Fig.4 TG curves of (a) pristine Al2Mo3O12 and (b) VTMS-g-Al2Mo3O12

It is considered that desorption of moisture takes place up to ~ 453 K (Figure 4),

while above this temperature the dehydration of primary bonded hydroxyl groups is the

predominant mass loss process [53]. On the other hand, the total mass loss of VTMS-g-

Al2Mo3O12 is ~ 0.5 mass % more than for pristine Al2Mo3O12 for the same temperature

interval (298 K 853 K), corroborating grafting of VTMS on Al2Mo3O12. It is worth

noting that up to 453 K (see Figure 4) the mass loss of VTMS-g-Al2Mo3O12 is lower

than for pristine Al2Mo3O12, indicating that the moisture content in VTMS-g-Al2Mo3O12

is lower due to its more hydrophobic nature.Furthermore, the mass loss between 453 K

and 853 K (~ 1.2 mass %) is predominantly due to the release of the VTMS fragments,

possibly vinyl groups, grafted on Al2Mo3O12 nanoparticles [53]. This could be indirect

evidence for primary bonding between Si from reactive silanol [53] and the surface of

Al2Mo3O12 through Si-O-Al and/or Si-O-Mo linkages. The successive processes of

hydrolysis, oligomerization and grafting, through the condensation of VTMS onto the

surface of Al2Mo3O12 nanoparticles, are presented in Figure 5.

13

Fig.5Scheme of (a) hydrolysis of VTMS; (b) oligomerization of reactive silanol

molecules; (c) formation of hydrogen bonds between silanol oligomers and Al2Mo3O12

nanoparticles and (d) grafting of VTMS on Al2Mo3O12 surface through primary bonds

3.4 VTMS-g-Al2Mo3O12/PE composites: Coefficient of thermal expansion, thermal

stability and mechanical properties

CTEs of neat PE and VTMS-g-Al2Mo3O12/PE composites were calculated from

the slopes of the respective thermal strain curves in the temperature range between 303

K and 373 K (Table 1).

Table 1. Thermal properties and degree of crystallinity of VTMS-g-Al2Mo3O12/PE composites with different filler contents. Nominal filler contents are presented; TGA shows inhomogeneity of the order of 0.5 mass %, regarding to nominal contents, for the samples used in dilatometry. CTE values are standard deviations based on triplicate measurements.

Filler content(mass

%) CTE (x10-4 K-1)

Mass loss (%) up to

573 K

Mass loss (%) up to

673 K T10(K)

T50 (K)

Tm (K) Xc (%)

0.0 2.290 0.0005 0.90 1.95 722 749 414 76 0.5 1.98 0.26 0.07 0.88 728 751 415 73 1.0 2.12 0.31 0.13 1.02 726 750 414 70 2.0 1.80 0.29 0.02 0.84 727 751 415 73 3.0 1.78 0.04 0.0 0.91 721 749 412 74 4.0 1.23 0.11 0.0 0.53 729 752 413 75

14

Large reductions of CTE, in comparison to neat PE, have been shown for almost

all VTMS-g-Al2Mo3O12/PE blends (0.5 - 4 mass %of filler), reaching a reduction of ~

46% for a loading of 4 mass %. The observed dependence of the average experimental

CTEs (Table 1) and volume fraction of VTMS-g-Al2Mo3O12 filler (Figure 6) can be

successfully described by Turners model[15]:

, (2)

where Km and Kf are the bulk moduli of PE (matrix) and Al2Mo3O12 (filler), is the

volume fraction of filler, and c, mand f are theCTEs of the composite, PE (matrix)

and Al2Mo3O12 (filler), respectively. Other models, such as the rule of mixtures, and

those of Schapery, Vo or Kerner, have not been able to model the experimental

dependence on average CTEs of the composites on volume fraction of filler.

Fig.6Average experimental CTEs of VTMS-g-Al2Mo3O12/PE composites, with standard

deviation bars, as a function ofnominal volume fraction of filler(black squares) fitted by

Turners model (red curve). Standard deviation bar for neat PE is smaller than the

symbol.

The literature valueof f = 9.5 x 10-6 K-1was considered for monoclinic

Al2Mo3O12[9], while m (PE) was taken from Table 1 (m = 2.29 x 10-4 K-1). The bulk

modulus of PE (Km) was estimated as 0.278 GPafrom the mean value of Youngs

modulus (Em = 0.278 GPa) measured for neat PE (Figure 7), assuming:

, (3)

15

where m is shear modulus of matrix, and

. (4)

Considering these values it is possible, using the Turner model, to determine the

bulk modulus of monoclinic Al2Mo3O12 (Kf) as 13.6 2.6 GPa, which agrees with the

ultrasonic results within the margin of error (cf. 3.2), validating the use of Turners

model.

Thermal stabilities of neat PE and VTMS-g-Al2Mo3O12/PE composites have

been evaluated from mass loss curves through parameters such as mass loss at 573 K,

mass loss at 673 K, temperature corresponding to 10% mass loss (T10) and temperature

corresponding to 50% mass loss (T50), as summarized in Table 1. Although the mass

losses of neat PE at 573 and 673 K are low, 0.90 and 1.95 mass % respectively, the

mass losses for all VTMS-g-Al2Mo3O12/PE compositesare far lower. T10 and

T50parameters of VTMS-g-Al2Mo3O12/PE composites are higher than for neat PE,

however, the filler impact on these two parameters was not as significant as observed

for the mass losses until 673 K. On the other hand, the effect of VTMS-g-Al2Mo3O12 filler on melting temperature (Tm) of the composites, as evaluated by DSC curves, was

negligible (Table 1). To evaluate the effect of VTMS-g-Al2Mo3O12 filler on degree of

crystallinity (Xc) of PE matrix, enthalpy change of melting (H) was determined for all

composites and their degrees of crystallinity were calculated considering H* = 273Jg-1

for completely crystallized PE [54]using:

(5)

where is the mass fraction of filler.

This evaluation showed that VTMS-g-Al2Mo3O12/PE composites lost some crystallinity,

from ~76% for neat PE to ~70% for VTMS-g-Al2Mo3O12/PE composite reinforced with

1 mass % of filler.

Youngs moduli, yield strengths and yield strains of composites were determined

from stress-strain curves and plotted in Figure 7. The stiffness and resistance to plastic

deformation of the newcomposites both increased, as evidenced by an increase in

Youngs modulus, between 6.6 and 19.1 %, and yield strength, between 4 and 6 %. On

the other hand, yield strain decreased between 8 and 12 % in comparison to neat PE.

16

Fig.7 (a) Youngs modulus; (b) Strength and strain at yield point of VTMS-g-Al2Mo-

3O12/PE composites as a function ofnominal filler content

FEAwas used to calculate the thermal stresses experienced in the composites

following a 100 K temperature drop, near room temperature. Cooling of the composites

induces compressive thermal stresses in the VTMS-g-Al2Mo3O12 and tensile thermal

stresses in the PE matrix due to their mismatched CTEs. The compressive stress in the

filler particles reached a maximum of 22 MPa, and was invariant with the filler loading.

The maximum tensile and shear stresses in the PE matrix are shown in Figure 8. The

maximum tensile stress increased approximately linearly with filler loading, while the

maximum shear stress increasednonlinearly.

Fig.8 Maximum tensile and shear stresses in the PE matrix as a function of filler

content. An estimated error due to finite mesh size of 1% is shown.

The CTE of the compositeon addition of filler was calculated to decrease only

slightly, but more than would be predicted by the rule of mixtures. The discrepancy

17

between the finite element models and the experimental result is possibly due to the

finite size of the model, which prevents accurate modelling of the constraint of the

thermal expansion of the matrix due to the filler since the bulk of the matrix material is

unconstrained.

4. Discussion

The mechanical properties of thermomiotics and low CTE materials still need to

be more thoroughly studied, sincethis knowledge is fundamental for many potential

applications. Thus far elastic constants and related properties such as bulk modulus (K)

have been determined for several compounds from the A2M3O12 family, in both

orthorhombic and monoclinic structures[55-58]. Our analysis of experimental c versus

volume filler fraction in terms of Turners model (Figure 6) and calculationsbased on

the ultrasonic testsgave a bulk modulus of 13.6 2.6 GPa and 13.5 1.3 GPa,

respectively, for monoclinic Al2Mo3O12, very close to the bulk moduli of monoclinic

Sc2W3O12, Sc2Mo3O12, In2W3O12 and Zr2WO4(PO4)2 of 11.8 0.8 GPa, 16 1 GPa, 13

2 GPa and 17 2 GPa, respectively, as determined from high pressure X-ray or neutron

diffraction data.Only the bulk modulus of monoclinic Al2W3O12 was reported to be

higher (28 1 GPa)[56], compared to other monoclinic phases. It is relevant for our

study that Watanabe et al.,[18] estimated the bulk modulus of Zr2WO4(PO4)2 at 18 GPa,

independently from theBirch-Murnaghan equation of state analysis of high-pressure X-

ray diffraction data [57], and by applying Hashins model on the relationship between

experimental Zr2WO4(PO4)2/Mgcomposite bulk modulus and Zr2WO4(PO4)2 filler

volume fraction, with excellent agreement for the two approaches. Monoclinic phases of

the A2M3O12 family are much softer than the corresponding orthorhombic phases, the

ones that present NTE, withbulk moduli varying between 32 GPa for Sc2W3O12 andSc2Mo3O12, and 48 GPa to 49 GPa for Al2W3O12 and Zr2WO4(PO4)2,

respectively[55-58]. In comparison, the bulk modulus of cubic -ZrW2O8, the most

studied thermomiotic filler, is much higher (~ 74 GPa)[59]. Although monoclinic

A2M3O12 phases have surprisingly low bulk moduli, even significantly lower than

already soft NTE orthorhombic phases, note that, for example, Al2Mo3O12 possesses a

bulk modulus two orders of magnitude higher than polyethyleneused in this study,

estimatedto be ~ 0.278 GPa. Therefore, in accordance to the Turner model, which

18

considers perfect interfaces between the two components [60], the high relative bulk

modulus of Al2Mo3O12 in comparison to the bulk modulus of PE would lead to a

relevant reduction of composite CTE. Strong reduction of the CTE of the composite (as

high as 46%) reached using filler contents not higher than 4 mass % (1.1 vol %) is a

possible indication of efficient filler-matrix interaction through physical entanglements.

It is worth noting that such a high reduction of composite CTE,obtained with low filler

volume fractions, is reached in spite of the small decrease in crystallinity of the PE

matrix with addition of VTMS-g-Al2Mo3O12fillers (Table 1), indicating that the fillers

thermal and mechanical properties, as well as the interfaces, dictate the overall

composite properties.

These findings underline the importance of well-dispersed nanoparticles in the

matrix as well as achievement of strong filler-matrix interactions, since the fulfillment

of these conditions results in increased density of well-bonded interfaces in the

composite per unit of volume, when compared to traditional non-functionalized micron

size fillers. For instance, Dey and Tripathi [36] reported the influence of Si micron

sized, chemically untreated, powder on thermal properties of Si/PE composites.

Although the CTEof the Si/PE composite decreased with increased volume percentage

of Si, the effect of filler content, as high as 20 vol %, on CTE of the composite was

lower than that reached in VTMS-g-Al2Mo3O12/PE composites with a volume filler

content of 1.1 vol % (4 mass %). In agreement with the Si/PE result[36], Anjana et al.,

[38]achieved similar reduction of CTE to that reached here for 1.1 vol % VTMS-g-

Al2Mo3O12/PE composite by adding 30 vol % of non-functionalized micron size CeO2 to the PE matrix. On the other hand, Sahebian et al.[34]reached a more pronounced

decrease of CTE by adding to PE matrix a lower percentage (10 vol %) of nanometric

CaCO3 that had been pre-treated with stearic acid. They also observed that the variation

of mean size of CaCO3 nanoparticles mayplay a role in the CTE of CaCO3/PE

composites; i.e., nanopowders with smaller mean particle size act more efficiently to

decrease the CTE of a composite than do nanopowders with higher mean particle size.

However, it is relevant to notice that Sharma et al.[25] inferred that the CTE of

ZrW2O8/polyimide composites was independent of the size of ZrW2O8 filler. They

showed that the 5 vol % loading of nanometric or micronic ZrW2O8 fillers to polyimide

(PI) matrix resulted in comparable CTE values. Sullivan and Lukehart[15] came to the

same conclusion for the effect of 0.8 and 1.7 vol % loadings of nanometric and micronic

ZrW2O8 fillers in a polyimide matrix. Sharma et al.[25]did not observe any additional

19

effect of ZrW2O8 nanofillers functionalized with 3-aminopropyltriethoxysilane or

ZrW2O8 fillers functionalized with 3-aminopropyltriethoxysilane/PI oligomer on

reduction of the composite CTE for loadings between 5 and 15 vol % when compared to

non-functionalized ZrW2O8 nanofillers. Therefore, those authors suggested that a

mechanism different from that usually ascribed for enhancement of mechanical and

thermal properties of nanocomposites, such as good filler-matrix interaction, is

responsible for governing thermal expansion in nanocomposites.

The thermal stability of VTMS-g-Al2Mo3O12/PE composites was significantly

increased at lower temperatures (< 673 K,Table 1), probably due to the joint action of

(i) low permeability of fillers leading to a more tortuous path of diffused species,

thereby decreasing the rate of gas evolution, and (ii) reduced movement of polymer

chains due to the interactions between VTMS-g-Al2Mo3O12 and the PE matrix [61].

Neat PE lost ~ 0.9 mass % on heating to 573 K, while the corresponding VTMS-g-

Al2Mo3O12/PE composites were essentially completely stable up to this temperature.

Further temperature increases up to 673 K provoked a total mass loss of neat PE of ~ 2

mass %, while mass losses of VTMS-g-Al2Mo3O12 composites were kept between 0.5

mass % (for 4 mass % VTMS-g-Al2Mo3O12) and ~ 1 mass % (for1 mass % of VTMS-g-

Al2Mo3O12). In comparison, Chrissafis et al.[62]reported a mass loss of 1 mass % up to

548 K for cSiO2 /PE composite with 2.5 mass % of filler (cSiO2: functionalized SiO2

nanofillers). Chrissafis and Bikiaris[61]reported that the addition of 2.5 mass % of

montmorillonite nanoclays (MMT) or multiwalled carbon nanotubes (MWCNT) to PE

matrix resulted in the same level of mass loss as their neat PE up to 573 K, 0.5 mass

%.

The origin of thermal stability of nanocomposites with polymer matrices at high

temperatures is not clear-cut since discrepant results have been reported, including

decreases of thermal stability, especially for nanoclay fillers, due to the activation of

different thermal degradation promoter mechanisms[61]. In the case of PE athigh

temperatures (> 723 K), where the PE matrix degradation rate is high, a small increase

of stability is confirmed, i.e., T10 and T50 were at most 6K and 3K higher, respectively,

for VTMS-g-Al2Mo3O12/PE composites with respect to neat PE. Here the increase of

thermal stability at high temperatures (>723 K) is very similar to the value (~ 4K)

observed by Chrissafis et al.[62]for a cSiO2 /PE composite with 2.5 mass % of filler,

judging from the temperature at which PE decomposition rate is highest. Similarly, a

20

rise of T10 of ~ 10 K was identified for the loadings < 5 mass % of non-functionalized

micron-sized ZrW2O8 in PI matrix [21]. However, much higher increases in T10 and T50,

25K and 21K, respectively, with respect to neat PE, were reported by Liao and

Tjong[49]for 4 mass % VTMS-g-Al2O3/PE composite. Therefore, it seems that low or

negative CTE fillers, like the majority of other inorganic fillers, do not promote thermal

degradation of polyolefins or polyimide matrices, and are capable of increasing their

thermal stability at high temperatures. In comparison, the thermal stability at high

temperatures of silane-crosslinked PE could increase as much as 35 Kas reported by

Aizan and Rahman[63]. Still, it is worth noting that in the future it would be relevant to

understand the effect of nanofillers onevolution of VOCs from PE matrices modified by

nanofillers, since the use ofPE for water pipe applications is negatively affected by the

migration of VOCs to drinking water.

The degree of crystallinity of PE was little reduced with increasing content of

VTMS-g-Al2Mo3O12nanofillers (Table 1), from 76% (neat PE) to 70% (1mass %

VTMS-g-Al2Mo3O12/PE composite).Although it is commonly expected that nanofillers

act only as heterogeneous nucleation sites, increasing the crystalline content of PE

matrix, a steric hindrance effect of nanofillers on chain ordering, resulting in reduction

of the crystallinity degree of PE has been reported[62, 64-66]. The hindrance effect of

MWCNT[65], pristine and organofilized MMT [62], nano-SiO2 and functionalized

nano-SiO2[62]and teak wood flour [66]on crystallization of PE is explained by the

obstruction of fillers in PE chain ordering. Our results suggest that the same steric

hindrance mechanism is responsible for the small reduction of the degree of crystallinity

in thePE matrix when VTMS-g-Al2Mo3O12 is added.The steric hindrance mechanism is

actually a well-known tool in polymer processing, used to inhibit crystallinity of

elastomers through addition of bulky methyl groups, rather than nanoparticles as here.

Also, silane crosslinking of PE, using coupling agents such as VTMS, reduces the

crystallinity of crosslinked PE due to steric hindrance[63]. Therefore, functionalized

nanoparticles could serve as defect centers impeding chain ordering and, thus, reduce

crystallinity of the PE matrix.

Melting temperature (Tm) was insensitive to the addition of nanofillers in

polymeric matrices. Our findings showed negligible variation of melting temperatures

of VTMS-g-Al2Mo3O12/PE composites compared with Tmof neat PE (Table 1). Many

other authors noticed asimilar lack of response of Tmin different composites with PE

matrix and nanofillers. The list of fillers added to PE matrix that do not affect melting

21

temperature of composite is extensive. For example, Chrissafis et al. [62]reported this

behavior for MWCNT, pristine and organofilized MMT, nano-SiO2 and functionalized

nano-SiO2 , Li et al. [65]for MWCNT, Liao and Tjong [49]for Al2O3 and VTMS-g-

Al2O3, Kanagaraj et al.[67]for MWCNT and Zebarjad et al. [68]for CaCO3. The

previously discussed decrease of crystallinity of the PE matrix in VTMS-g-

Al2Mo3O12/PE composites could reduce the sizes of crystal lamellae inPE spherulites.

On the other hand, it is known that reduction of the average sizes of crystal lamellae

decreases Tm of polymers. Therefore, it can be rationalized that negligible variation of

Tm in VTMS-g-Al2Mo3O12/PE composites is possibly due to the simultaneous action of

two opposing effects: (i) size reduction of crystal lamellae inside spherulites causing a

decrease of Tm and (ii) efficient filler-matrix interaction through physical entanglements

contributing to an increase of Tm.

The mechanical properties of VTMS-g-Al2Mo3O12/PE composites, such as

Youngs modulus and yield strength, were improved at room temperature in comparison

to neat PE (Figure 7). The increase of Youngs modulus (between 6.6 to 19.1 %), and

yield strength (between 4 and 6 %) are a consequence of much higher elastic constants

of monoclinic Al2Mo3O12 in comparison to PE, but principally arises from the efficient

load transfer from the PE matrix to the filler owing to the high density of strong filler-

matrix interfaces, reached through physical entanglements. It is worth noting that the

overall Youngs modulus and yield strength of VTMS-g-Al2Mo3O12/PE composites

increased despite the reduction in the degree of crystallinity of PE matrixfor

thecomposites, since it is well established that the increase of amorphous content in PE

should lead to a decrease of mechanical properties of the matrix at the temperatures

above the glass transition temperature (~ 163 K for PE). However, in VTMS-g-

Al2Mo3O12/PE composites the intimate interfacespermit load transfer, overriding the

impact of decreasing crystallinity of PE matrix. On the other hand, there are opposite

examples in the literature such as the report of Li et al. [65]which attributed the

reduction of Youngs modulus of MWCNT/PE composites to the decrease of

crystallinity of PE matrix.

The reduction of yield strain,as observed for VTMS-g-Al2Mo3O12/PE

composites (Figure 7),has been reported for other composites with polymeric and

metallic matrices. For example, Sharma et al. [25]reported a strong decrease of yield

strain for 5 to 15 mass % additions inside PI matrix of ZrW2O8 nanofillers

functionalized with 3-aminopropyltriethoxysilane or with 3-

22

aminopropyltriethoxysilane/PI oligomer. It should be noticed that even the addition of

ductile Cu particles in a PE matrix [69]resultedin reduction of yield strain of Cu/PE

composites.

The thermal stresses calculated by the finite element method for VTMS-g-

Al2Mo3O12/PE composites following cooling by 100 K(Fig. 8) were below the yield

strength of the PE matrix (Fig. 7b) as well as the level that would be expected to affect

the Al2Mo3O12filler[9]. Therefore, the composites would be expected to be

mechanically stable following thermal shock of 100 K or less, due to the absence of

plastic deformation of the matrix or fracture of the Al2Mo3O12 filler.However, addition

of Al2Mo3O12 in amounts considerably above 1.1 vol % might increase the thermal

stress in the matrix above the yield strength of the PE due to the nonlinear increase in

the thermal shear stress in the matrix. Although there is considerable CTE mismatch

between PE and Al2Mo3O12, the low stiffness of the matrix reduces the magnitude of the

thermal stresses considerably, in addition to allowing reduction of the CTE by addition

of only a small amount of filler.

5. Conclusions

This study investigated the viability of employinga nanometric A2M3O12-

typecompound, in low loadings ( 4 mass %), as a filler in a thermoplastic polymer in

order to achieve considerable reduction of the matrix CTE, without causing adverse

effects on themechanical properties and thermal stability. Nanometric Al2Mo3O12, in the

monoclinic form, grafted with VTMS, abi-functional coupling agent, was added to a PE

matrix through melt-compounding extrusion and VTMS-g-Al2Mo3O12/PE composites

were fabricated by injection molding.

The as-prepared VTMS-g-Al2Mo3O12/PE composites exhibited strong decrease

of their CTEs, reaching 46 % reduction for loadings of 4 mass % (1.1 vol %). The

results suggest efficient filler-matrix interaction through physical entanglements.

Experimentally obtained CTEs of VTMS-g-Al2Mo3O12/PE composites were

successfully described by Turners model and the bulk modulus of monoclinic

Al2Mo3O12 was calculated to be 13.6 2.6 GPa, in good agreement with the bulk

modulus of 13.5 1.3GPa obtained by ultrasonic testing.

Thermal stability of VTMS-g-Al2Mo3O12/PE composites was significantly

increased at lower temperatures (< 673 K) in comparison to neat PE, while at high

temperatures the increase was small, but not negligible. The degree of crystallinity of

PE decreased a small extent with VTMS-g-Al2Mo3O12 loading. However, the increased

23

amount of amorphous PE did not result in a significant change in melting temperature

or deterioration of mechanical properties, the latter as judged by the Youngs modulus

and yield strength. Actually, the Youngs modulus increased by 6.6 to 19.1 %, while the

yield strength increased 4to 6 % owing to efficient load transfer from PE matrix to the

fillers, because of the high density of strong filler-matrix interfaces.

This research shows the importance of further studies of low or negative

CTE/thermoplastic composites.

Acknowledgments

B.A. Marinkovic and J.R.M. dAlmeida are grateful to CNPq (National Council

for Scientific and Technological Development) for a Research Productivity Grants.

Patricia I. Pontn is also grateful to CNPq for scholarship.M.A. White acknowledges

support of NSERC through the Discovery Grants program. We thank J.W. Zwanziger

for use of the ultrasonic transducer.

References

[1] Romao CP, Miller KJ, Whitman CA, White MA, Marinkovic BA(2013) Negative

Thermal Expansion (Thermomiotic) Materials. In: Reedijk J, Poeppelmeier K(eds)

Comprehensive Inorganic Chemistry II, vol. 4.Elsevier, Oxford, pp 127-151

[2] EvansJS, Mary TA, Vogt T, Subramanian MA, Sleight AW (1996) Negative

thermal expansion in ZrW2O8 and HfW2O8,Chem Mater8:2809-2823.

[3] Evans J S, Mary TA, Sleight AW (1997)Negative thermal expansion in a large

molybdate and tungstate family, J Solid State Chem133:580-583.

[4] Lind C, Wilkinson AP, Hu Z, Short S, Jorgensen JD(1998) Synthesis and properties

of the negative thermal expansion material cubic ZrMo2O8,Chem Mater 10:2335-

2337.

[5] EvansJS, Mary TA, Sleight AW (1998) Negative thermal expansion in Sc2(WO4)3,

J Solid State Chem 137:148-160.

[6] Evans JS, Mary TA (2000) Structural phase transitions and negative thermal

expansion in Sc2(MoO4)3, Int J Inorg Mater 2:143-151.

[7] Sumitra S, Waghmare UV, Umarji AM (2007) Anomalous dynamical charges,

phonons, and the origin of negative thermal expansion in Y2W3O12 Phys. Rev. B

76:024307.

24

[8] MarinkovicBA, Ari M, de Avillez RR, Rizzo F, FerreiraFF, Miller KJ, Johnson M

B, White MA (2009) Correlation between AO6polyhedral distortion and negative

thermal expansion in orthorhombic Y2Mo3O12 and related materials,

ChemMater21:2886-2894.

[9] Prisco LP, Romao CP, Rizzo F, White MA, Marinkovic BA (2013) The effect of

microstructure on thermal expansion coefficients in powder-processed Al2Mo3O12,

J Mater Sci 48:2986-2996.

[10] KimIJ andGauckler LJ (2008) Excellent thermal shock resistant materials with low

thermal expansion coefficients,J Ceram Process Res 9:240-245.

[11] Shi JD, Pu ZJ, Wu KH, Larkins G (1997) Composite materials with adjustable

thermal expansion for electronic applications,Mater Res Soc Symp Proc 445:229-

234.

[12] Holzer H and Dunand DC (1999) Phase transformation and thermal expansion of

Cu/ZrW2O8 metal matrix composites, J Mater Res14:780-789.

[13] Kofteros M, Rodriguez S, Tandon V, Murr LE (2001) A preliminary study of

thermal expansion compensation in cement by ZrW2O8 additions, Scripta Mater

45:369-374.

[14] Tran KD, Groshens TJ, Nelson JG (2001) Fabrication of near-zero thermal

expansion (FexSc1-X)2Mo3O12-MoO3 ceramic composite using the reaction sintering

process, Mater Sci and Eng A 303:234-240.

[15] Sullivan LM and Lukehart CM (2005) Zirconium tungstate (ZrW2O8)/Polyimide

nanocomposites exhibiting reduced coefficient of thermal expansion,

ChemMater17:2136-2141.

[16] Tani J, Kimura H, Hirota K, Kido H (2007) Thermal expansion and mechanical

properties of phenolic resin/ZrW2O8composites, JAppl Polym Sci106:3343-3347.

[17] Goertzen WKand Kessler MR (2008) Three-phase cyanate ester composites with

fumed silica and negative-CTE reinforcements, JTherm Anal Calorim 93:87-93.

[18] Watanabe H, Tani J, Kido H, Mizzuchi K (2008) Thermal expansion and

mechanical propertis of pure magnesim containing zirconium tungsten phosphate

particles with negative thermal expansion, Mater Sci Eng A 494:291-298.

[19] Yanase I, Miyagi M, Kobayashi H (2009) Fabrication of zero-thermal-expansion

ZrSiO4/Y2W3O12 sintered body, J Eur Ceram Soc29:3129-134.

[20] Kanamori K, Kineri T, Fukuda R, Kawano T, Nishio K (2009) Low-temperature

sintering of ZrW2O8-SiO2 by spark plasma sintering, JMater Sci 44:855-869.

25

[21] Yang J, Yang Y, Qinqin L, Guifang X, Cheng X (2010) Preparation of negative

thermal expansion ZrW2O8powders and its application in polyimide/ZrW2O8

composites, J Mater Sci Technol 26:665-668.

[22] Tani J, Takahashi M, Kido H (2010) Fabrication and thermal expansion properties

of ZrW2O8/Zr2WP2O12 composites, J Eur Ceram Soc 30:1483-1488.

[23] Lind C, Coleman MR, Kozy LC, Sharma GR (2011) Zirconium tungstate/polymer

nanocomposites: challenges and opportuniteis, Phys Status Solidi B248:123-129.

[24] Chu X, Huang R, Yang H, Wu Z, Lu J, Zhou Y, Li L (2011) The cryogenic thermal

expansion and mechanical properties of plasma modified ZrW2O8 reinforced

epoxy,Mater Sci Eng A 528:3367-3374.

[25] Sharma GR, Lind C, Coleman MR (2012) Preparation and properties of polyimide

nanocomposites with negative thermal expansion nanoparticle filler,

MaterChemPhys137:448-457.

[26] YamashinaN, Isobe T, Ando S (2012) Low thermal expansion composites prepared

from polyimide and ZrW2O8 particles with negative thermal expasion,JPhotopolym

Sci Technol25:385-388.

[27] Wu Y, Wang M, Chen Z, Ma N,Wang H (2013) The effect of phase transformation

on the thermal expansion property in Al/ZrW2O8 composites, JMater Sci48:2928-

2933.

[28] Peng Z, Sun YZ, Peng LM (2014) Hydrothermal synthesis of ZrW2O8 nanorods

and its application in ZrW2O8/Cu composites with controllable thermal expansion

coefficients, MateDes54:989-994.

[29] Liu QQ, Cheng XN, Yang J, Sun XJ (2011) Influence of fabrication method on the

structure and thermal expansion property of ZrWMoO8 and its composites, JMater

Sci 46:1253-1258.

[30] Suzhu Y, Hing P, Hu X (2000) Thermal expansion behaviour of polystyrene-

aluminium nitride composites, J Phys D: Appl Phys 33:1606-1610.

[31] Wong CP and Bollampally RS (1999) Thermal conductivity, elastic modulus, and

coefficient of thermal expansion of polymer composites filled with ceramic

particles for electronic packaging,JApplPolym Sci 74:3396-3403.

[32] Take WA, Watson E, Brachman RW, Rowe RK (2012) Thermal expansion and

contraction of geomembrane liners subjected to solar exposure and backfilling,

JGeotech Geoenviron138:1287-1397.

26

[33] Skjevrak I, Due A, Gjerstad KO, Herikstad H (2003) Volatile organic components

migrating from plastic pipes (HDPE, PEX and PVC) into drink water, Water

Res37:1912-1920.

[34] Sahebian S, Zebarjad SM, Sajjadi SA (2010) Role of surface active agent on

dimensional stability of HDPE/CaCO3nanocomposites, J Thermoplast

ComposMater 23:583-596.

[35] Baglari S, Kole M, Dey TK (2011) Effective thermal conductivity and coefficient

of linear thermal expansion of high-density polyethylene - fly ash composites,

Indian JPhys85:559-573.

[36]Dey T K and Tripathi M (2010) Thermal properties of silicon powder filled high-

density polyethylene composites, Thermochim Acta502:35-42.

[37]Manu KM, Ananthakumar S, Sebastian MT (2013) Electrical and thermal

properties of low permittivity Sr2Al2SiO7 ceramic filled HDPE composites,

CeramInt39:4945-4951.

[38]Anjana PS, Deepu V, Uma S, Mohanan P, Philip J, Sebastian MT (2010) Dielectric,

thermal, and mechanical properties of CeO2-Filled HDPE composites for

microwave substrate applications, J Polym Sci, Part B: Polym Phys48:998-1008.

[39]Pllnen M, Suihkonen R, Nevalainen K, Koistinen AP, Suvanto M, Vuorinen J,

Pakkanen T (2012) Morphological, mechanical, tribological, and thermal expansion

properties of organoclay reinforced polyethylene composites, Polym EngSci

53:1279-1286.

[40]Liu QQ, Cheng XN, Yang J (2012) Development of low thermal expansion

Sc2(WO4)3 containing composites, Mater Tech27:388-392.

[41]MarinkovicBA, Ari M, Jardim PM, de Avillez RR, Rizzo F, Ferreira FF (2010)

In2Mo3O12: A low negative thermal expansion compound,

ThermochimActa499:48-53.

[42]Xiao XL, Cheng YZ, Peng J, Wu MM, Chen DF, Hu ZB, Kiyanagi R, Fieramosca

JS, Short S, Jorgensen J (2008) Thermal expansion properties of A2(MO4)3 (A = Ho

and Tm; M = W and Mo), Solid State Sci 10:321-325.

[43]Ari M, Jardim PM, Marinkovic BA, Rizzo F, Ferreira FF (2008) Thermal expansion

of Cr2xFe22xMo3O12, Al2xFe22xMo3O12 and Al2xCr22xMo3O12 solid solutions, JSolid

State Chem181:1472-1479.

27

[44]Ari M, Miller KJ, Marinkovic BA, Jardim PM, de Avillez R, Rizzo F, White MA

(2011) Rapid synthesis of the low thermal expansion phase of Al2Mo3O12 via a sol

gel method using polyvinyl alcohol, JSol-Gel Sci Technol58:121-125.

[45]Pontn PI, d'Almeida JM, Marinkovic BA, Savic SM, Mancic L, Rey NA, Morgado

EJr, Rizzo FC (2014) The effects of the chemical composition of titanate nanotubes

and solvent type on 3-aminopropyltriethoxysilane grafting efficiency, Appl Surf Sci

301:315-322.

[46]Asmani M, Kermel C, Leriche A, Ourak M (2001) Influence of porosity on Young's

modulus and Poisson ration in alumina ceramics, J Eur Ceram Soc21:1081-1086.

[47] http://www.paralab.pt/sites/default/files/pdf/DIL402C.pdf. Accessed 17 Feb 2014

[48]Kuznetsova A, Wovchko EA, Yates JT (1997) FTIR study of the adsorption and

thermal behavior of vinyltriethoxysilane chemisorbed on -Al2O3,

Langmuir13:5322-5328.

[49]Liao CZ and Tjong SC (2013) Mechanical and thermal perfomance of high-density

polyethylene/alunima nanocomposites, J Macromol Sci, Part B: Phys52:812-825.

[50]Nguyen VG, Thai H, Mai HD, Tran HT, Tran DL, Vu M T (2013) Effect of

titanium dioxide on the properties of polyethylene/TiO2 nanocomposites,

Composites Part B45:1192-1198.

[51]Abboud M, Turner M, Duguet E, Fontanille M (1997) PMMA-based composite

materials with reactive ceramic fillers Part 1.Chemical modification and

characterisation of ceramic particles, J Mater Chem7:1527-1532.

[52]Byrne MT, McCarthy EJ, Bent M, Blake R (2007) Chemical functionalisation of

titania nanotubes and their utilisation for the fabrication of reinforced polystyrene

composites, JMater Chem 17:2351-2358.

[53]Guo Z, Pereira T, Choi O, Wang Y, Hahn HT (2006) Surface functionalized

alumina nanoparticle filled polymeric nanocomposites with enhanced mechanical

properties, J Mater Chem16:2800-2808.

[54]Gao J-g, Yu M-s, Li Z-t (2004) Nonisothermal crystallization kinetics and melting

behavior of bimodal medium density polyethylene/low density polyethylene blends,

Eur Polym J44:15331539.

[55]Varga T, Wilkinsin AP, Jorgensen JD, Short S (2006) Neutron powder diffraction

study of the orthorhombic to monoclinic transition in Sc2W3O12 on compression,

Solid State Sci 8:289-295.

28

[56]Varga T, Wilkinson AP, Lind C, Bassett WA, Zha C (2005) High pressure

synchrotron x-ray powder diffraction study of Sc2Mo3O12 and Al2W3O12,

JPhys:CondensMatter17:4271-4283.

[57]Baiz TI, Heinrich PC, Banek NA, Vivekens BL, Lind C (2012) In-situ non-ambient

X-ray diffraction studies of indium tungstate, JSolid State Chem 187:195-199.

[58]Cetinkol M, Wilkinsion AP, Lind C (2009) In situ high-pressure synchrotron x-ray

diffraction study of Zr2(WO4)(PO4)2 up to 16 GPa, PhysRev B79:224118.

[59]Drymiotis FR, Ledbetter H, Betts JB, Kimura T, Lashley JC, Migliori A (2004)

Monocrystal elastic constants of the negative-thermal-expansion compound

zirconium tungstate (ZrW2O8), Phys Rev Lett93:025502.

[60]Takenaka K (2012) Negative thermal expansion materials: technological key for

control of thermal expansion, Sci Technol Adv Mater 13:013001.

[61]Chrissafis K and BikiarisD (2011) Can nanoparticles really enhance thermal

stability of polymers? Part I: An overview on thermal decomposition of addition

polymers,Thermochim Acta, 523:1-24.

[62]Chrissafis K, Paraskevopoulos KM, Tsiaoussis I, Bikiaris D (2009) Comparative

study of the effect of different nanoparticles on the mechanical properties,

permeability, and thermal degradation mechanism of HDPE, JAppl PolymSci

114:1606-1618.

[63]Aizan W and Rahman WA (2006) Design of silane crosslinkable high density

polyethylene compounds for automotive fuel tank application. Proyect Report,

Universiti Teknologi Malaysia.

[64]Tjong SC (2006) Structural and mechanical properties of polymer nanocomposites,

Mater Sci Eng R53:73-197.

[65]Li S, Chen H, Cui D, Li J, Zhang Z, Wang Y, Tang T (2010) Structure and

properties of multi-walled carbon nanotubes/polyethylene nanocomposites

synthesized by in situ polymerization with supported Cp2ZrCl2 catalyst, Polym

Compos 31:507-515.

[66]Sewda K and Maiti SN (2009) Mechanical properties of teak wood flour-reinforced

HDPE composites, J ApplPolym Sci112:1826-1834.

[67]Kanagaraj S, Varanda FR, Zhil'tsova TV, Oliveira MS, Simes JAO (2007)

Mechanical properties of high density polyethylene/carbon nanotube composites,

Compos SciTechnol 67:3071-3077.

29

[68]Zebarjad SM, Sajjadi SA, Tahani M, Lazzeri A (2006) A study on thermal

behaviourof HDPE/CaCO3 nanocomposites, J Achiev Mater Manuf Eng 17:173-

176.

[69]Tavman IH (1997) Thermal and mechanical properties of copper powder filled

poly(ethylene) composites, Powder Technol 91:63-67.